Reader Jamie at the research department of the University of California asks an interesting question about the predictive capability of ISM data.

Hi Mish,

I’m wondering what a 3-mo rolling average of the ISM reports would look like, plotted against the quarterly GDP report of annualized growth. If there is any predictive value in the ISM reports, I’d expect GDP to roll over into negative territory, and the Q2 number to be revised down a bit. How much noise is there in the ISM? Does these reports work as coincident indicators with any degree of accuracy? Just thinking this might make for an interesting post.


Thanks Jamie, let’s take a look.

Manufacturing ISM Noise

click on chart for sharper image

There are many false signals on manufacturing ISM, where a contraction does not imply a recession. Other manufacturing ISM charts show the same noise.

Let’s turn our attention to the services ISM. Unfortunately, the series only dates to 1997.

Service ISM New Orders

click on chart for sharper image

Service ISM Business Activity

click on chart for sharper image

14 years is not a lot of data. Moreover, there is even less data for the index than the components, at least on FRED where I picked up the charts. However, the pattern for the data we do have is somewhat clear.

A dip in 2003 did not precede a recession, but that dip was short-lived.

There are 73 charts in the ISM series so inquiring minds may wish to take a closer look.

I commented on the services ISM Business Activity number on Wednesday in ISM says “Business Conditions Flattening Out”; Why Services Number Worse Than It Looks; Unsustainable Conditions.

Unsustainable Conditions

Production [business activity] is +2.7 while new orders, employment, and deliveries are down. Also note that backlog of orders has plunged over the past two months. Meanwhile new export orders is not only in a free-fall, but also in contraction for the first month as the global economy cools.

Supplier deliveries are on the verge of contraction, and inventories were +3 points to 56.5.

In short, one of these numbers does not make sense in relation to the others, in relation to the manufacturing ISM, in relation to the financial industry, and in relation to the global economy.

That 56.1 production reading at +2.7 simply does not fit in, and is not sustainable if the other conditions remain in current “slowing” condition.

The possibility of a much bigger decline next month seems very real. In fact, that is my call in advance.

Real GDP Percent Change From Year Ago

There is plenty of recession predictive or coincident capability in “Real” GDP (inflation adjusted GDP)

click on chart for sharper image

Nearly every time Real GDP dips below 2%, the economy was either in recession or headed for recession.

Services ISM confirms as do many other data points including consumer spending and jobs. This chart suggests we are headed for recession if not already in one.

More Than Meets the Eye

I wrote the above several days ago. There was so much other immediate news that I never got around to publishing it. Since then I read an article by John Hussman essentially saying essentially same thing.

Please give Hussman’s post More Than Meets the Eye a well-deserved look.

The components of our recession warning composite might be called “weak learners” in that none of them, individually, has a particularly notable record in anticipating recessions. The full syndrome of conditions, however, captures a critical “signature” of recessions. That signature of “early warning” conditions is based on financial market indicators including credit spreads, equity prices and yield curve behavior, coupled with slowing in measures of employment and business activity. Every historical instance of this full syndrome has been associated with an ongoing or immediately impending recession.

The components (which I’ve reordered for simplicity) are:

1: Widening credit spreads: An increase over the past 6 months in either the spread between commercial paper and 3-month Treasury yields, or between the Dow Corporate Bond Index yield and 10-year Treasury yields.

2: Falling stock prices: S&P; 500 below its level of 6 months earlier. This is not terribly unusual by itself, which is why people say that market declines have called 11 of the past 6 recessions, but falling stock prices are very important as part of the broader syndrome.

3: Weak ISM Purchasing Managers Index: PMI below 50, or,

3: (alternate): Moderating ISM and employment growth: PMI below 54, coupled with slowing employment growth: either total nonfarm employment growth below 1.3% over the preceding year (this is a figure that Marty Zweig noted in a Barron’s piece many years ago), or an unemployment rate up 0.4% or more from its 12-month low.

4: Moderate or flat yield curve: 10-year Treasury yield no more than 2.5% above 3-month Treasury yields if condition 3 is in effect, or any difference of less than 3.1% if 3(alternate) is in effect (again, this criterion doesn’t create a strong risk of recession in and of itself).

At present, both measures of credit spreads in condition 1 are widening, the S&P; 500 is within about one percent of its level 6 months ago, the Purchasing Managers Index is at 55.3%, total nonfarm payrolls have grown by only 0.8% over the past year, the unemployment rate is up 0.4% from its March 2011 low, and the Treasury yield spread is just 2.7%. From the standpoint of this composite, we would require only modest deterioration in stock prices and the ISM index to produce serious recession concerns.

Wells Fargo’s senior economist Mark Vitner reiterated the point last week, noting that since 1950, year-over-year growth in real GDP has dipped below 2% on 12 occasions. In 10 of those instances, the economy was already in recession or quickly entered one. The exceptions were 1956 and 2003.

For our part, we’ve always believed that the strongest evidence is obtained by combining multiple data points into a single “gestalt.” So I have difficulty concluding that the U.S. is on the verge of recession simply because the year-over-year growth rate has stalled. At the same time, we are closely monitoring a much broader set of data, because the deterioration has been very rapid. I should be clear – the evidence is not yet convincing that a recession is imminent, but it is also important to recognize that the developing risks are greater than most investors seem to assume at present.

Definition of Recession

The NBER, which is the official arbiter of recessions describes recessions this way.

The NBER does not define a recession in terms of two consecutive quarters of decline in real GDP. Rather, a recession is a significant decline in economic activity spread across the economy, lasting more than a few months, normally visible in real GDP, real income, employment, industrial production, and wholesale-retail sales.

“Sufficient” Does Not Mean “Necessary”

Two declining quarters of GDP is a “sufficient” recession condition, however, not even one quarter of declining real GDP is a “necessary” condition.

The recession that started in November of 2007 did not have one full quarter of declining Real GDP growth.

Real GDP

Notice that chart shows declining “growth”. GDP was actually rising at the time the recession started.

Real GDP Percent Change vs. Year Ago

Those waiting for contraction before they concede the US is in recession may wake up one day and discover, just as happened in 2008, that the recession was 1/3 over before they saw it “coming”. Indeed, some recessions may not be spotted until they are already over.

Might the US be in recession now?

One thing is for sure: At a minimum, the US is certainly on the border of one. Economist Dave Rosenberg raised his odds of recession this week from 99% to 100%. That is how certain he is.

For the average guy on the street out of work and unable to find any job, the last recession never ended.

Mike “Mish” Shedlock
Click Here To Scroll Thru My Recent Post List